1. Field of the Invention
The present invention relates generally to the field of optical imaging and more particularly to catadioptric optical systems used for bright-field and dark-field imaging applications.
2. Description of the Related Art
Many optical and electronic systems exist to inspect object surfaces for defects such as those on a partially fabricated integrated circuit or a photomask. Defects may take the form of particles randomly localized on the surface of the circuit or photomask, as well as scratches, process variations, and so forth. Various imaging techniques used to perform surface inspection for such defects provide different advantages depending on the types of defects present.
Two well known imaging techniques for detecting defects are bright field imaging and dark field imaging. Bright field imaging is commonly used in microscope systems. The advantage of bright field imaging is the image produced is readily distinguishable. The size of image features accurately represents the size of object features multiplied by the magnification of the optical system. Bright field imaging can be more easily used with image comparison and processing algorithms for computerized object comparison, defect detection, and classification.
Dark field imaging has been successfully used to detect irregularities on object surfaces. The advantage of dark field imaging is that flat specular areas scatter very little light toward the detector, resulting in a dark image. Any surface features or objects protruding above the object scatter light toward the detector. Thus, in inspecting objects like semiconductor wafers, dark field imaging produces an image of features, particles, or other irregularities on a dark background.
One advantage of dark field imaging is that it provides a large signal for small features that scatter light. This large signal allows dark field imaging to detect smaller object features and provide faster object inspections than bright field imaging. Another advantage is that Fourier filtering can be used to minimize repeating pattern signals and enhance the defect signal-to-noise ratio.
Many dark field imaging techniques have been developed to enhance the detection of different types of defects. These techniques consist of a specific illumination scheme and collection scheme such that the scattered and diffracted light collected from the object provides the best signal. Several optical systems have been developed that use different dark field imaging techniques including laser directional dark field, double dark field, central dark ground, and ring dark field.
When employing either bright field or dark field imaging it is often desirable to use short wavelength illumination in the 300-400 nm ultraviolet (UV) range, 200-300 nm deep ultraviolet (deep UV or DUV) range, or 100-200 vacuum ultraviolet (vacuum UV or VUV) range. For bright field imaging short wavelength illumination provides improved resolution allowing the detection of smaller object features. For dark field imaging short wavelength illumination provides greatly increased scattering signals that allow the detection of smaller objects, an increase in inspection speed, or a decrease in the illumination power requirements. In addition, both bright field and dark field imaging can take advantage of changes in material absorption and reflectivities at short wavelengths. The changes in absorption and reflectivity of different materials at short wavelengths can help to identify these different materials. Also, many materials have greatly increased absorption at wavelengths in the DUV and VUV. Increased absorption can help improve optical inspection of upper surfaces, such as in semiconductor wafer inspection, by minimizing reflections interference from underlying layers.
Optical systems supporting bright field and dark field imaging typically require correction over some finite spectral bandwidth or wavelength range. Correction is necessary because different wavelengths have different glass indexes, known as dispersion. Conventional designs usually use two or three glass types to compensate for dispersive effects. Compensating for these dispersive effects is called color correction. Color correction in the UV, DUV, and VUV wavelength ranges is increasingly difficult. At shorter wavelengths, the glass dispersion greatly increases and is difficult to correct. In addition, at shorter wavelengths fewer and fewer glass materials may be used for correction.
At wavelengths shorter than 365 nm there are very few glass materials having high transmission. These materials typically include silica, CaF2, MgF2, and LiF2. Of these materials, silica is most desirable to use in high end optical systems. Silica is a hard glass with low thermal expansion, no birefringence, high UV damage threshold, and is not sensitive to humidity. CaF2, MgF2, and LiF2 are soft glasses which are difficult to polish, have high thermal expansion, some birefringence, and can be sensitive to humidity. Of these fluoride glasses, CaF2 is the most desirable to use as an optical glass.
Minimizing the number of glass materials used in a UV, DUV or VUV optical system produces special challenges for correcting color aberrations. This is especially true in the VUV wavelength range where both silica and CaF2 are extremely dispersive. Even a narrow spectral bandwidth at very short wavelengths can require the correction of numerous distinct color aberrations. Some important color aberrations that need to be corrected may include primary and secondary axial color, primary and secondary lateral color, chromatic variation of spherical aberration, and chromatic variation of coma.
At a wavelength of 157 nm, for example, CaF2 is the only reasonable glass material that has high transmission and does not have severe problems with birefringence, water solubility, or mechanical softness. Standard color correction is not possible because no other glass material is available.
Another problem with currently available systems is that such systems provide a relatively short working distance between the optical system and the surface being inspected. Photomask inspection requires the working distance of the imaging system to be greater than approximately 6 millimeters due to the protective pellicle present on the photomask. A long working distance is also desirable in laser dark-field inspection environments. An imaging system having a long working distance makes it possible to directly illuminate the surface being inspected from outside the objective. Under typical circumstances, a working distance greater than 4 millimeters presents generally desirable attributes, while a working distance greater than 8 millimeters is preferred.
Further, a high numerical aperture (NA) provides advantages for high resolution imaging and collecting as large a solid angle as possible. It is highly desirable to achieve numerical apertures of 0.8, which corresponds to collecting angles above the surface from normal to 53 degrees.
A further problem with currently available systems is that while some relatively high NA systems exist, the central returning rays may be obscured due to apertures and other optical components. Such a central obscuration blocks low frequency information from the image and is undesirable.
Additionally, some presently available systems include internal pupil planes. A system having an internal pupil plane is undesirable because it does not readily support aperturing, particularly variations in aperture shapes, and Fourier filtering.
Finally, some currently available systems have limited field sizes. A large field size is often important for area imaging and to allow high speed inspection such as for semiconductor wafers and photomasks. Field size is typically limited by aberrations such as lateral color and chromatic variation of aberrations. Aberration correction is especially difficult if combined with chromatic correction for a large spectral bandwidth, high NA, long working distances, no central obscuration, and an external pupil plane.
Two prior patents describe high NA catadioptric systems that can support this type of imaging. These patents are U.S. Pat. No. 5,717,518 to Shafer et al. and U.S. patent application Ser. No. 09/046,814 to Chuang et al, now U.S. Pat. No. 6,064,517, both assigned to KLA-Tencor Corporation and hereby fully incorporated by reference.
U.S. Pat. No. 5,717,518 describes an apparatus capable of high NA, ultra broad-band UV imaging. The '518 patent presents a 0.9 NA system for broad-band bright field and multiple wavelength dark-field imaging. The '518 system has a high degree of chromatic correction using a single glass material. Further correction is possible using two glass materials. The '518 system employs an achromatized field lens group to correct for secondary and higher order lateral color. This design has several limitations, including a limited working distance, central obscuration, internal pupil, and relatively tight manufacturing tolerances. The primary method for laser dark-field illumination in this scheme is to direct a laser through a hole or aperture in the spherical mirror element. This type of illumination can be quite complicated to implement.
U.S. patent application Ser. No. 09/046,814, now U.S. Pat. No. 6,064,517 discloses an apparatus capable of combining ultra high NA, narrow-band UV imaging and multiple laser dark-field imaging techniques into a single optical system. The design is single wavelength and operates at numerical apertures up to 0.99. The [′814 application] ′517 patent is ideally suited for use in laser dark-field inspection, but has several limitations, including a relatively narrow working distance, a central obscuration, a narrow bandwidth, and relatively tight manufacturing tolerances. The system of the [′814 application] ′517 patent uses a similar technique to that shown in the ′ 518 patent for laser dark-field illumination and generally has similar advantages and limitations.
Other specialized catadioptric optical systems have been developed for use in semiconductor lithography. These systems are designed to image a photomask at a reduced magnification onto a resist coated wafer. Two prior patents describe high NA catadioptric systems that can support this type of imaging, specifically U.S. Pat. No. 5,052,763 to Singh et al. and European Patent Application number EP 0 736 789 A2 to Takahashi.
The '763 patent describes a catadioptric optical system capable of high NA imaging. This optical system is designed to create a substantially flat image field over the large areas required for semiconductor lithography. The design utilizes an input optical system having a curved field, a catadioptric relay system, and an output optical system to correct for the field curvature and some monochromatic aberrations. Limitations for the '763 design include a limited working distance, an internal pupil, a narrow bandwidth, an internal beamsplitter, and tight manufacturing tolerances.
The Takahashi European Patent Application presents a catadioptric optical system capable of high NA imaging. This optical system is designed to reduce the required diameter of the catadioptric mirror element for long distances to the wafer. The Takahashi design has an internal pupil, a narrow bandwidth using multiple glass materials, and tight manufacturing tolerances.
It is therefore desirable to provide a system for performing both bright field and dark field surface inspection having an objective which corrects image aberrations, chromatic variation of image aberrations, and longitudinal (axial) color and lateral color, including residual (secondary and higher order) lateral color correction over a broad spectral range. Such a system should be relatively inexpensive and easy to operate in typical environments, provide a relatively long working distance, large filed size, and have lenient tolerances. It is preferable to have such a system which operates at UV, DUV, or VUV wavelengths, a high numerical aperture, without central obscuration, and with an accessible pupil plane.